Bootstrap Procedure

The bootstrap procedure was proposed by Effron in order to estimate the standard errors of a sample for those statistics and for those sampling distributions where a conventional formula could not be used. It is increasingly being employed for complex and highly stratified samples, and for statistics whose sampling distribution is not known. The procedure involves multiple sampling at random with replacement from among the primary sampling units, to obtain the specified number of primary sampling units and a very large distribution of estimated values. The mean and standard deviation of the estimated values of the distribution so obtained are taken as the mean and standard deviation respectively of the estimated statistic.

Compiled by John P. Keeves 1995